At P10, hippocampal genesis is almost completed and the typical lamination is finished. Perikarya of the stratum pyramidal display the characteristic pyramidal shape 18. The immunofluorescent detection of CDK4 demonstrated a neuronal expression at P10 (Fig. 1A–F). Expression of CDK4 was localized both in nuclei and cytoplasm of pyramidal cells of CA1 regions. In addition, mossy fibers were immuno-reactive as well (data not shown). At P28. The sub-cellular localization of CDK4 was different. Staining of CDK4 could be detected predominantly in the cytoplasm but only sporadically in neuronal nuclei of pyramidal neurons of the CA subfields (Fig. 1G–L). Further quantitative immunoblotting analysis confirmed the similar expression levels of CDK4 in P10 and P28 mice hippocampus (Fig. 1M).

Having localized CDK4 in the hippocampus, we next determined the role of cyclinD1-CDK4 in synaptic plasticity. We first examined whether the complex were required for normal basal synaptic transmission. We compared extracellular recordings of the SC-CA1 synapses in hippocampal slices (n = 12 slices from 4 animals) to slices pretreated with CDK4 inhibitor (n = 16 slices from 5 animals). There was not significantly different both at neonatal and adolescent animals (P = 0.897 and 0.826 respectively, analysis of variance (ANOVA); Fig. 2A, B). Furthermore, the fEPSP slope corre-sponding to a given presynaptic fiber volley was also unaffected by CDK4 inhibitor application throughout development (P = 0.883 and 0.437, respectively, analysis of variance (ANOVA); Fig. 2C, D). Bath application of 5 μM CDK4 inhibitor to slices from P8-15 and P21-35 animals did not affect the fEPSP slope, even after 1 h in the perfusion buffer (n = 16 slices from 5mice; Fig. 2E, F). Previous studies have shown that these concentrations are adequate to inhibit CDK4 kinase activity in cells 26.

We next examined whether the complex were involved in paired-pulse facilitation, a form of short-term plasticity. As shown in Fig. 3, at six different inter-stimulus intervals ranging from 30 to 300 ms, the mean paired-pulse facilitation ratio (second fEPSP slope/first fEPSP slope) was reduced in CDK4 inhibitor pre-incubated slices (n = 11 slices from 3 mice) as compared with control (n = 10 slices from 3 mice), at time intervals from 30 to 200 ms (p < 0.01, Fig. 3A), at the neonatal stage (P8-15), surprisingly there was no significant decrease in PPF when applied to slices from adolescent animals (P21-35) (n = 12 slices from 4 mice respectively; P = 0.217, Fig. 3B). These results demonstrated that cyclinD1-CDK4 may have the effect of modulating pre-synaptic function at the neonatal stage but not at the adolescent stage. This difference may be the consequence of the expression and translocation of cyclinD1-CDK4 in neuronal cells in area CA1 at different development stage 18.

As a second measure of short-term synaptic plasticity, we next compared the time evolution of the fEPSP slopes during and following a brief tetanic stimulation. The Schaffer collaterals were stimulated by a short 40 Hz train (8 stimuli) followed by a test stimulus delivered 300 ms after the end of the burst. We found that in the CDK4 inhibitor pretreated slices facilitation was greatly reduced during the burst (stimuli 2, 3, 4, 5, 6, 7)(n = 12, 11 slices from 4 mice for P8-15, P21-35; P < 0.01 and 0.005, respectively; Fig. 3C and 3D) compared with control slices (n = 11, 10 slices from 3mice for P8-15, P21-35, respectively) and while the 300 ms after the burst was not significantly different (ANOVA, P = 0. 86 and 0.37 compared with control, respectively). These observations suggested that cyclinD1-CDK4 may regulate the availability of vesicles from the readily releasable pool during repetitive stimulation during the postnatal development.

To examine the role of cylinD1-CDK4 in long-term synaptic plasticity, we tested LTP at the SC-CA1 synapses. LTP was induced by HFS (100-Hz, 1s) to the Schaffer collateral inputs. The amount of LTP was quantified as the average fEPSP slope 45 min after HFS relative to the baseline mean slope. Inhibition of CDK4 had no significant effect on the induction or maintenance of LTP after 45 min both for P8-15 (control 163 ± 3%, 13 slices from 5 mice; control + CDK4I 164 ± 6%, 11 slices from 5 mice; Fig. 4A) and P21-35 (control 162 ± 8%, 13 slices from 6mice; control + CDK4I 167 ± 6%, 11 slices from 5mice; Fig. 4B). We did not examine whether the expression of this form of LTP or other forms of LTP, such as those induced by theta stimulation or non-NMDA receptor-dependent LTP in the CA3 region, are equally unaffected by CDK4 inhibitor.

A metabotropic glutamate receptor-dependent form of LTD (mGluR-LTD) is expressed more strongly in neonatal synapses and can be preferentially induced by DHPG, a selective group I mGluR agonist. Therefore, the ability to develop and maintain this form of LTD was tested by applying DHPG (100 μM) to hippocampal slices. As shown in Fig. 4, CDK4 inhibitor did not alter the acute depression of synaptic transmission induced by DHPG, in contrast, 10 min after washout of DHPG, the fEPSP slope increased gradually in CDK4 inhibitor pretreated slices, while in control slices it remained depressed. In control slices, the fEPSP slope was reduced to 79 ± 2.8% of the baseline average value 45 min after application of DHPG (n = 15 slices from 5 mice), whereas, preincubation of slices with the CDK4 inhibitor, significantly inhibited the later phase of DHPG-LTD, fEPSP slope had recovered to 95 ± 4.1% of average baseline (n = 12 slices from 6 mice; ANOVA, p < 0.001) (Fig. 5A) in neonatal animals (P8 -P15). DHPG-LTD was also reduced by CDK4 inhibitor in adolescent animals (P21-P35) compared with control slices (DHPG plus vehicle, 78 ± 3% of the baseline, n = 15 slices from 5 mice; DHPG plus CDK4 inhibitor, 89 ± 3.6%, n = 12 slices from 6 mice, ANOVA, p < 0.01) (Fig. 5B).

When LTD was induced by PP-LFS (paired pulse separated by 50 ms were delivered at low frequency 1 Hz for 15 min in the presence of the NMDAR antagonist D-AP-5), the results on LTD induction and maintenance were similar to those obtained with DHPG-induced LTD. The later phase of the PP-LFS-induced LTD almost disappeared by CDK4 inhibitor (fEPSP slope returned to 96 ± 4%, n = 12 slices from 5 mice) in the P8-15 group while the fEPSP slope in the P21-35 group (n= 11 slices from 5mice) recovered to 91 ± 3.2% of the baseline at 45 min (ANOVA, p < 0.001 and 0.01, respectively) (Fig. 5C and 5D).

A different form of LTD, NMDA receptor dependent LTD (NMDAR-LTD), has been described at CA1 synapses. Consistent with the idea that mGluR- and NMDAR-LTD are mediated by distinct cellular mechanisms 2728, CDK4 inhibitor- treated slices showed similar levels of NMDAR-LTD induced by LFS (1 Hz, 15min, 900 pulses) to control slices both from neonatal and adolescent animals as shown in Fig. 5E, F. After 45 min, fEPSP slope was 78 ± 2.2% and 81 ± 2.4% in slices from P8-15 and P21-35 mice respectively, moreover 79 ± 2.5% and 82 ± 3.6% in CDK4 inhibitor pretreated slices. These results indicated that cyclinD1-CDK4 are specifically required for mGluR dependent LTD, but do not play a significant role in NMDAR-LTD.

Care of animals and experiments were conducted in accordance with the National Institutes of Health Guideline for the Care and Use of Laboratory Animals. Efforts were made to minimize the number of animals used. All mice were of the C57B6/129 genetic background. Most of the experiments were conducted in the brain slices of mice at postnatal day8 (P8) to 15 (P15), and day21 (P21) to 35 (P35). The number of mice or hippocampal slices used in each individual experiment was indicated as the "n" value in the figure legend.

Neonatal or adolescent mice were perfused with ice-cold saline followed by fixative containing 4% PFA in 0.1 M PBS, pH 7.4. Brains were removed, postfixed in the same solution over night, and subsequently cryoprotected by equilibration with 30% sucrose in PBS. Sections of 30 μm thickness were cut on freezing microtome and collected in 0.1 M TBS, pH 7.4. Antigen unmasking was performed by treating sections with 0.05% trypsin for 30 min at room temperature. After rinsing in PBS, sections were blocked with 5%bovine serum albumin (BSA) in PBS with 0.3% Triton X-100 (PBST + BSA) for 30 min. Afterwards, sections were incubated with anti-CDK4 antibody (1:300, Santa Cruz Biotechnology) overnight at 4°C. Section-bound antibodies were revealed by incubation with biotinylated goat anti-rabbit antiserum (1:1000, 1h, Amersham). All incubation steps were separated by intense washing with TBS. Specificity of labeling was tested by omitting primary antisera. After such incubation, tissue was devoid of immunoreactivity as expected. Images were captured with the digital microscope camera AxioCamHRC running on Axiovision 3.1 Software (both Carl Zeiss Jena) and processed by means of Adobe1 Photoshop1 7.0 (Adobe Systems, Mountain View, CA). Sections were nuclear counter-stained with DAPI.

P8-15 and P21-35 mice were decapitated and the brain was quickly removed and immersed in ice-cold artificial cerebrospinal fluid (ACSF) saturated with 95% O₂/5% CO₂ and containing 124.0 mM NaCl, 3.0 mM KCl, 1.25 mM KH₂PO₄, 26.0 mM NaHCO₃, 2.0 mM MgSO₄, 2.5 mM CaCl₂, and 10.0 mM glucose. Hippocampi were isolated and cut into 400 μm thick transversal slices by a custom-made tissue-slicer. Slices were maintained in ACSF at room temperature for at least 1 h before recording.

Slices were transferred to a submerged recording chamber, held submerged between two nylon nets, maintained at 30°C, and perfused continuously with ACSF at a rate of 2ml-3ml/min. Field excitatory postsynaptic potentials (fEPSPs) were recorded with glass microelectrodess (1–3 MΩ) filled with ACSF and positioned in stratum radiatum of area CA1. Synaptic responses were reliably evoked every 20sec in the CA1 region of the hippocampus by stimulating Schaffer collaterals with 0.1 ms pulses. A full input-output (I/O) curve was constructed at the beginning of each experiment, the stimulus intensity (0.2 ms duration, 0.033 Hz) selected for baseline measurement was adjusted to elicit about 40% of the maximal slope. Paired pulse facilitation (PPF) was tested by applying two pulses at different intervals (30–300 ms). LTP was induced using high frequency stimulation (HFS; 100 Hz for 1 s), after 20 min of stable baseline recordings. mGluR-LTD was induced by application of 100 μM DHPG (Tocris Cookson) for 5 min or by pairs of stimuli (interstimulus interval, 50 ms) delivered at 1Hz for 15 min (1800pulses; PPF-LFS).

D, L-AP-5 (Sigma) was prepared fresh in ACSF. CDK4 inhibitor (Merck-Calbiochem)(2-Bromo-12,13-dihydro-5H-indolo [2,3-a]pyrrolo [3,4-c]carbazole-5,7(6H)-dione) was dissolved in dimethylsulphoxide (DMSO) at a concentration of 5 μM. DHPG (Tocris Cookson) was prepared as a 1000 × stock in DMSO, used fresh or kept frozen at -20°C. In both cases, the final concentration of DMSO was at most 0.1% v/v after dilution in the recording bath. Slices were pre-incubated with the inhibitors for 20–30 min before DHPG or paired-pulse low-frequency stimulation (PP-LFS). The effects of all of the pharmacological treatments on LTD were evaluated by comparing control and treated slices.

Normalized data were averaged across experiments and expressed as means ± SEM. The PPF ratio was obtained by dividing the field potential slope from the second pulse (fEPSP2) by that of the first pulse (fEPSP1). One-way ANOVA was performed to determine whether there were significant differences followed by Bonferroni test as post hoc analysis; p < 0.01 indicated difference significance. Origin analysis software (Microcal Software, Northampton, MA) was used in data analysis.